Add complete MNMD Docker deployment example with startup coordination and VolumeNotFound fix (#642)

* Initial plan

* Add MNMD Docker deployment example with 4 nodes x 4 drives

- Create docs/examples/mnmd/ directory structure
- Add docker-compose.yml with proper disk indexing (1..4)
- Add wait-and-start.sh for startup coordination
- Add README.md with usage instructions and alternatives
- Add CHECKLIST.md with step-by-step verification
- Fixes VolumeNotFound issue by using correct volume paths
- Implements health checks and startup ordering
- Uses service names for stable inter-node addressing

Co-authored-by: houseme <4829346+houseme@users.noreply.github.com>

* Add docs/examples README as index for deployment examples

Co-authored-by: houseme <4829346+houseme@users.noreply.github.com>

* Add automated test script for MNMD deployment

- Add test-deployment.sh with comprehensive validation
- Test container status, health, endpoints, connectivity
- Update README to reference test script
- Make script executable

Co-authored-by: houseme <4829346+houseme@users.noreply.github.com>

* improve code

* improve code

* improve dep crates `cargo shear --fix`

* upgrade aws-sdk-s3

---------

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: houseme <4829346+houseme@users.noreply.github.com>
Co-authored-by: houseme <housemecn@gmail.com>
This commit is contained in:
Copilot
2025-10-12 13:15:14 +08:00
committed by GitHub
parent aac9b1edb7
commit ad99019749
10 changed files with 1050 additions and 107 deletions

60
docs/examples/README.md Normal file
View File

@@ -0,0 +1,60 @@
# RustFS Deployment Examples
This directory contains practical deployment examples and configurations for RustFS.
## Available Examples
### [MNMD (Multi-Node Multi-Drive)](./mnmd/)
Complete Docker Compose example for deploying RustFS in a 4-node, 4-drive-per-node configuration.
**Features:**
- Proper disk indexing (1..4) to avoid VolumeNotFound errors
- Startup coordination via `wait-and-start.sh` script
- Service discovery using Docker service names
- Health checks with alternatives for different base images
- Comprehensive documentation and verification checklist
**Use Case:** Production-ready multi-node deployment for high availability and performance.
**Quick Start:**
```bash
cd docs/examples/mnmd
docker-compose up -d
```
**See also:**
- [MNMD README](./mnmd/README.md) - Detailed usage guide
- [MNMD CHECKLIST](./mnmd/CHECKLIST.md) - Step-by-step verification
## Other Deployment Examples
For additional deployment examples, see:
- [`examples/`](/examples/) - Root-level examples directory with:
- `docker-quickstart.sh` - Quick start script for basic deployments
- `enhanced-docker-deployment.sh` - Advanced deployment scenarios
- `docker-comprehensive.yml` - Docker Compose with multiple profiles
- [`.docker/compose/`](/.docker/compose/) - Docker Compose configurations:
- `docker-compose.cluster.yaml` - Basic cluster setup
- `docker-compose.observability.yaml` - Observability stack integration
## Related Documentation
- [Console & Endpoint Service Separation](../console-separation.md)
- [Environment Variables](../ENVIRONMENT_VARIABLES.md)
- [Performance Testing](../PERFORMANCE_TESTING.md)
## Contributing
When adding new examples:
1. Create a dedicated subdirectory under `docs/examples/`
2. Include a comprehensive README.md
3. Provide working configuration files
4. Add verification steps or checklists
5. Document common issues and troubleshooting
## Support
For issues or questions:
- GitHub Issues: https://github.com/rustfs/rustfs/issues
- Documentation: https://rustfs.io

View File

@@ -0,0 +1,329 @@
# MNMD Deployment Checklist
This checklist provides step-by-step verification for deploying RustFS in MNMD (Multi-Node Multi-Drive) mode using
Docker.
## Pre-Deployment Checks
### 1. System Requirements
- [ ] Docker Engine 20.10+ installed
- [ ] Docker Compose 2.0+ installed
- [ ] At least 8GB RAM available
- [ ] At least 40GB disk space available (for 4 nodes × 4 volumes)
Verify with:
```bash
docker --version
docker-compose --version
free -h
df -h
```
### 2. File System Checks
- [ ] Using XFS, ext4, or another suitable filesystem (not NFS for production)
- [ ] File system supports extended attributes
Verify with:
```bash
df -T | grep -E '(xfs|ext4)'
```
### 3. Permissions and SELinux
- [ ] Current user is in `docker` group or can run `sudo docker`
- [ ] SELinux is properly configured (if enabled)
Verify with:
```bash
groups | grep docker
getenforce # If enabled, should show "Permissive" or "Enforcing" with proper policies
```
### 4. Network Configuration
- [ ] Ports 9000-9031 are available
- [ ] No firewall blocking Docker bridge network
Verify with:
```bash
# Check if ports are free
netstat -tuln | grep -E ':(9000|9001|9010|9011|9020|9021|9030|9031)'
# Should return nothing if ports are free
```
### 5. Files Present
- [ ] `docker-compose.yml` exists in current directory
Verify with:
```bash
cd docs/examples/mnmd
ls -la
chmod +x wait-and-start.sh # If needed
```
## Deployment Steps
### 1. Start the Cluster
- [ ] Navigate to the example directory
- [ ] Pull the latest RustFS image
- [ ] Start the cluster
```bash
cd docs/examples/mnmd
docker-compose pull
docker-compose up -d
```
### 2. Monitor Startup
- [ ] Watch container logs during startup
- [ ] Verify no VolumeNotFound errors
- [ ] Check that peer discovery completes
```bash
# Watch all logs
docker-compose logs -f
# Watch specific node
docker-compose logs -f rustfs-node1
# Look for successful startup messages
docker-compose logs | grep -i "ready\|listening\|started"
```
### 3. Verify Container Status
- [ ] All 4 containers are running
- [ ] All 4 containers show as healthy
```bash
docker-compose ps
# Expected output: 4 containers in "Up" state with "healthy" status
```
### 4. Check Health Endpoints
- [ ] API health endpoints respond on all nodes
- [ ] Console health endpoints respond on all nodes
```bash
# Test API endpoints
curl http://localhost:9000/health
curl http://localhost:9010/health
curl http://localhost:9020/health
curl http://localhost:9030/health
# Test Console endpoints
curl http://localhost:9001/health
curl http://localhost:9011/health
curl http://localhost:9021/health
curl http://localhost:9031/health
# All should return successful health status
```
## Post-Deployment Verification
### 1. In-Container Checks
- [ ] Data directories exist
- [ ] Directories have correct permissions
- [ ] RustFS process is running
```bash
# Check node1
docker exec rustfs-node1 ls -la /data/
docker exec rustfs-node1 ps aux | grep rustfs
# Verify all 4 data directories exist
docker exec rustfs-node1 ls -d /data/rustfs{1..4}
```
### 2. DNS and Network Validation
- [ ] Service names resolve correctly
- [ ] Inter-node connectivity works
```bash
# DNS resolution test
docker exec rustfs-node1 nslookup rustfs-node2
docker exec rustfs-node1 nslookup rustfs-node3
docker exec rustfs-node1 nslookup rustfs-node4
# Connectivity test (using nc if available)
docker exec rustfs-node1 nc -zv rustfs-node2 9000
docker exec rustfs-node1 nc -zv rustfs-node3 9000
docker exec rustfs-node1 nc -zv rustfs-node4 9000
# Or using telnet/curl
docker exec rustfs-node1 curl -v http://rustfs-node2:9000/health
```
### 3. Volume Configuration Validation
- [ ] RUSTFS_VOLUMES environment variable is correct
- [ ] All 16 endpoints are configured (4 nodes × 4 drives)
```bash
# Check environment variable
docker exec rustfs-node1 env | grep RUSTFS_VOLUMES
# Expected output:
# RUSTFS_VOLUMES=http://rustfs-node{1...4}:9000/data/rustfs{1...4}
```
### 4. Cluster Functionality
- [ ] Can list buckets via API
- [ ] Can create a bucket
- [ ] Can upload an object
- [ ] Can download an object
```bash
# Configure AWS CLI or s3cmd
export AWS_ACCESS_KEY_ID=rustfsadmin
export AWS_SECRET_ACCESS_KEY=rustfsadmin
# Using AWS CLI (if installed)
aws --endpoint-url http://localhost:9000 s3 mb s3://test-bucket
aws --endpoint-url http://localhost:9000 s3 ls
echo "test content" > test.txt
aws --endpoint-url http://localhost:9000 s3 cp test.txt s3://test-bucket/
aws --endpoint-url http://localhost:9000 s3 ls s3://test-bucket/
aws --endpoint-url http://localhost:9000 s3 cp s3://test-bucket/test.txt downloaded.txt
cat downloaded.txt
# Or using curl
curl -X PUT http://localhost:9000/test-bucket \
-H "Host: localhost:9000" \
--user rustfsadmin:rustfsadmin
```
### 5. Healthcheck Verification
- [ ] Docker reports all services as healthy
- [ ] Healthcheck scripts work in containers
```bash
# Check Docker health status
docker inspect rustfs-node1 --format='{{.State.Health.Status}}'
docker inspect rustfs-node2 --format='{{.State.Health.Status}}'
docker inspect rustfs-node3 --format='{{.State.Health.Status}}'
docker inspect rustfs-node4 --format='{{.State.Health.Status}}'
# All should return "healthy"
# Test healthcheck command manually
docker exec rustfs-node1 nc -z localhost 9000
echo $? # Should be 0
```
## Troubleshooting Checks
### If VolumeNotFound Error Occurs
- [ ] Verify volume indexing starts at 1, not 0
- [ ] Check that RUSTFS_VOLUMES matches mounted paths
- [ ] Ensure all /data/rustfs{1..4} directories exist
```bash
# Check mounted volumes
docker inspect rustfs-node1 | jq '.[].Mounts'
# Verify directories in container
docker exec rustfs-node1 ls -la /data/
```
### If Healthcheck Fails
- [ ] Check if `nc` is available in the image
- [ ] Try alternative healthcheck (curl/wget)
- [ ] Increase `start_period` in docker-compose.yml
```bash
# Check if nc is available
docker exec rustfs-node1 which nc
# Test healthcheck manually
docker exec rustfs-node1 nc -z localhost 9000
# Check logs for errors
docker-compose logs rustfs-node1 | grep -i error
```
### If Startup Takes Too Long
- [ ] Check peer discovery timeout in logs
- [ ] Verify network connectivity between nodes
- [ ] Consider increasing timeout in wait-and-start.sh
```bash
# Check startup logs
docker-compose logs rustfs-node1 | grep -i "waiting\|peer\|timeout"
# Check network
docker network inspect mnmd_rustfs-mnmd
```
### If Containers Crash or Restart
- [ ] Review container logs
- [ ] Check resource usage (CPU/Memory)
- [ ] Verify no port conflicts
```bash
# View last crash logs
docker-compose logs --tail=100 rustfs-node1
# Check resource usage
docker stats --no-stream
# Check restart count
docker-compose ps
```
## Cleanup Checklist
When done testing:
- [ ] Stop the cluster: `docker-compose down`
- [ ] Remove volumes (optional, destroys data): `docker-compose down -v`
- [ ] Clean up dangling images: `docker image prune`
- [ ] Verify ports are released: `netstat -tuln | grep -E ':(9000|9001|9010|9011|9020|9021|9030|9031)'`
## Production Deployment Additional Checks
Before deploying to production:
- [ ] Change default credentials (RUSTFS_ACCESS_KEY, RUSTFS_SECRET_KEY)
- [ ] Configure TLS certificates
- [ ] Set up proper logging and monitoring
- [ ] Configure backups for volumes
- [ ] Review and adjust resource limits
- [ ] Set up external load balancer (if needed)
- [ ] Document disaster recovery procedures
- [ ] Test failover scenarios
- [ ] Verify data persistence after container restart
## Summary
This checklist ensures:
- ✓ Correct disk indexing (1..4 instead of 0..3)
- ✓ Proper startup coordination via wait-and-start.sh
- ✓ Service discovery via Docker service names
- ✓ Health checks function correctly
- ✓ All 16 endpoints (4 nodes × 4 drives) are operational
- ✓ No VolumeNotFound errors occur
For more details, see [README.md](./README.md) in this directory.

View File

@@ -0,0 +1,268 @@
# RustFS MNMD (Multi-Node Multi-Drive) Docker Example
This directory contains a complete, ready-to-use MNMD deployment example for RustFS with 4 nodes and 4 drives per node (
4x4 configuration).
## Overview
This example addresses common deployment issues including:
- **VolumeNotFound errors** - Fixed by using correct disk indexing (`/data/rustfs{1...4}` instead of
`/data/rustfs{0...3}`)
- **Startup race conditions** - Solved with a simple `sleep` command in each service.
- **Service discovery** - Uses Docker service names (`rustfs-node{1..4}`) instead of hard-coded IPs
- **Health checks** - Implements proper health monitoring with `nc` (with alternatives documented)
## Quick Start
From this directory (`docs/examples/mnmd`), run:
```bash
# Start the cluster
docker-compose up -d
# Check the status
docker-compose ps
# View logs
docker-compose logs -f
# Test the deployment
curl http://localhost:9000/health
curl http://localhost:9001/health
# Run comprehensive tests
./test-deployment.sh
# Stop the cluster
docker-compose down
# Clean up volumes (WARNING: deletes all data)
docker-compose down -v
```
## Configuration Details
### Volume Configuration
The example uses the following volume configuration:
```bash
RUSTFS_VOLUMES=http://rustfs-node{1...4}:9000/data/rustfs{1...4}
```
This expands to 16 endpoints (4 nodes × 4 drives):
- Node 1: `/data/rustfs1`, `/data/rustfs2`, `/data/rustfs3`, `/data/rustfs4`
- Node 2: `/data/rustfs1`, `/data/rustfs2`, `/data/rustfs3`, `/data/rustfs4`
- Node 3: `/data/rustfs1`, `/data/rustfs2`, `/data/rustfs3`, `/data/rustfs4`
- Node 4: `/data/rustfs1`, `/data/rustfs2`, `/data/rustfs3`, `/data/rustfs4`
**Important:** Disk indexing starts at 1 to match the mounted paths (`/data/rustfs1..4`).
### Port Mappings
| Node | API Port | Console Port |
|-------|----------|--------------|
| node1 | 9000 | 9001 |
| node2 | 9010 | 9011 |
| node3 | 9020 | 9021 |
| node4 | 9030 | 9031 |
### Startup Coordination
To prevent race conditions during startup where nodes might not find each other, a simple `sleep 3` command is added to
each service's command. This provides a brief delay, allowing the network and other services to initialize before RustFS
starts. For more complex scenarios, a more robust health-check dependency or an external entrypoint script might be
required.
### Health Checks
Default health check using `nc` (netcat):
```yaml
healthcheck:
test: [ "CMD-SHELL", "nc -z localhost 9000 || exit 1" ]
interval: 10s
timeout: 5s
retries: 3
start_period: 30s
```
#### Alternative Health Checks
If your base image lacks `nc`, use one of these alternatives:
**Using curl:**
```yaml
healthcheck:
test: [ "CMD-SHELL", "curl -f http://localhost:9000/health || exit 1" ]
interval: 10s
timeout: 5s
retries: 3
start_period: 30s
```
**Using wget:**
```yaml
healthcheck:
test: [ "CMD-SHELL", "wget --no-verbose --tries=1 --spider http://localhost:9000/health || exit 1" ]
interval: 10s
timeout: 5s
retries: 3
start_period: 30s
```
### Brace Expansion Alternatives
If your Docker Compose runtime doesn't support brace expansion (`{1...4}`), replace with explicit endpoints:
```yaml
environment:
- RUSTFS_VOLUMES=http://rustfs-node1:9000/data/rustfs1,http://rustfs-node1:9000/data/rustfs2,http://rustfs-node1:9000/data/rustfs3,http://rustfs-node1:9000/data/rustfs4,http://rustfs-node2:9000/data/rustfs1,http://rustfs-node2:9000/data/rustfs2,http://rustfs-node2:9000/data/rustfs3,http://rustfs-node2:9000/data/rustfs4,http://rustfs-node3:9000/data/rustfs1,http://rustfs-node3:9000/data/rustfs2,http://rustfs-node3:9000/data/rustfs3,http://rustfs-node3:9000/data/rustfs4,http://rustfs-node4:9000/data/rustfs1,http://rustfs-node4:9000/data/rustfs2,http://rustfs-node4:9000/data/rustfs3,http://rustfs-node4:9000/data/rustfs4
```
## Using RUSTFS_CMD
The `RUSTFS_CMD` environment variable provides a fallback when no command is specified:
```yaml
environment:
- RUSTFS_CMD=rustfs # Default fallback command
```
This allows the entrypoint to execute the correct command when Docker doesn't provide one.
## Testing the Deployment
After starting the cluster, verify it's working:
### Automated Testing
Use the provided test script for comprehensive validation:
```bash
./test-deployment.sh
```
This script tests:
- Container status (4/4 running)
- Health checks (4/4 healthy)
- API endpoints (4 ports)
- Console endpoints (4 ports)
- Inter-node connectivity
- Data directory existence
### Manual Testing
For manual verification:
```bash
# 1. Check all containers are healthy
docker-compose ps
# 2. Test API endpoints
for port in 9000 9010 9020 9030; do
echo "Testing port $port..."
curl -s http://localhost:${port}/health | jq '.'
done
# 3. Test console endpoints
for port in 9001 9011 9021 9031; do
echo "Testing console port $port..."
curl -s http://localhost:${port}/health | jq '.'
done
# 4. Check inter-node connectivity
docker exec rustfs-node1 nc -zv rustfs-node2 9000
docker exec rustfs-node1 nc -zv rustfs-node3 9000
docker exec rustfs-node1 nc -zv rustfs-node4 9000
```
## Troubleshooting
### VolumeNotFound Error
**Symptom:** Error message about `/data/rustfs0` not found.
**Solution:** This example uses `/data/rustfs{1...4}` indexing to match the mounted Docker volumes. Ensure your
`RUSTFS_VOLUMES` configuration starts at index 1, not 0.
### Health Check Failures
**Symptom:** Containers show as unhealthy.
**Solutions:**
1. Check if `nc` is available: `docker exec rustfs-node1 which nc`
2. Use alternative health checks (curl/wget) as documented above
3. Increase `start_period` if nodes need more time to initialize
### Startup Timeouts
**Symptom:** Services timeout waiting for peers.
**Solutions:**
1. Check logs: `docker-compose logs rustfs-node1`
2. Verify network connectivity: `docker-compose exec rustfs-node1 ping rustfs-node2`
3. Consider increasing the `sleep` duration in the `docker-compose.yml` `command` directive if a longer delay is needed.
### Permission Issues
**Symptom:** Cannot create directories or write data.
**Solution:** Ensure volumes have correct permissions or set `RUSTFS_UID` and `RUSTFS_GID` environment variables.
## Advanced Configuration
### Custom Credentials
Replace default credentials in production:
```yaml
environment:
- RUSTFS_ACCESS_KEY=your_access_key
- RUSTFS_SECRET_KEY=your_secret_key
```
### TLS Configuration
Add TLS certificates:
```yaml
volumes:
- ./certs:/opt/tls:ro
environment:
- RUSTFS_TLS_PATH=/opt/tls
```
### Resource Limits
Add resource constraints:
```yaml
deploy:
resources:
limits:
cpus: '2'
memory: 4G
reservations:
cpus: '1'
memory: 2G
```
## See Also
- [CHECKLIST.md](./CHECKLIST.md) - Step-by-step verification guide
- [../../console-separation.md](../../console-separation.md) - Console & endpoint service separation guide
- [../../../examples/docker-comprehensive.yml](../../../examples/docker-comprehensive.yml) - More deployment examples
- [Issue #618](https://github.com/rustfs/rustfs/issues/618) - Original VolumeNotFound issue
## References
- RustFS Documentation: https://rustfs.io
- Docker Compose Documentation: https://docs.docker.com/compose/

View File

@@ -0,0 +1,182 @@
# Copyright 2024 RustFS Team
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# MNMD (Multi-Node Multi-Drive) Docker Compose Example
# 4 nodes x 4 drives configuration
# This example demonstrates a complete, ready-to-use MNMD deployment
# addressing startup coordination and VolumeNotFound issues.
services:
rustfs-node1:
image: rustfs/rustfs:latest
container_name: rustfs-node1
hostname: rustfs-node1
environment:
# Use service names and correct disk indexing (1..4 to match mounted paths)
- RUSTFS_VOLUMES=http://rustfs-node{1...4}:9000/data/rustfs{1...4}
- RUSTFS_ADDRESS=0.0.0.0:9000
- RUSTFS_CONSOLE_ENABLE=true
- RUSTFS_CONSOLE_ADDRESS=0.0.0.0:9001
- RUSTFS_ACCESS_KEY=rustfsadmin
- RUSTFS_SECRET_KEY=rustfsadmin
- RUSTFS_CMD=rustfs
ports:
- "9000:9000" # API endpoint
- "9001:9001" # Console
volumes:
- node1-data1:/data/rustfs1
- node1-data2:/data/rustfs2
- node1-data3:/data/rustfs3
- node1-data4:/data/rustfs4
command: [ "sh", "-c", "sleep 3 && rustfs" ]
healthcheck:
test:
[
"CMD",
"sh", "-c",
"curl -f http://localhost:9000/health && curl -f http://localhost:9001/health"
]
interval: 10s
timeout: 5s
retries: 3
start_period: 30s
networks:
- rustfs-mnmd
rustfs-node2:
image: rustfs/rustfs:latest
container_name: rustfs-node2
hostname: rustfs-node2
environment:
- RUSTFS_VOLUMES=http://rustfs-node{1...4}:9000/data/rustfs{1...4}
- RUSTFS_ADDRESS=0.0.0.0:9000
- RUSTFS_CONSOLE_ENABLE=true
- RUSTFS_CONSOLE_ADDRESS=0.0.0.0:9001
- RUSTFS_ACCESS_KEY=rustfsadmin
- RUSTFS_SECRET_KEY=rustfsadmin
- RUSTFS_CMD=rustfs
ports:
- "9010:9000" # API endpoint
- "9011:9001" # Console
volumes:
- node2-data1:/data/rustfs1
- node2-data2:/data/rustfs2
- node2-data3:/data/rustfs3
- node2-data4:/data/rustfs4
command: [ "sh", "-c", "sleep 3 && rustfs" ]
healthcheck:
test:
[
"CMD",
"sh", "-c",
"curl -f http://localhost:9000/health && curl -f http://localhost:9001/health"
]
interval: 10s
timeout: 5s
retries: 3
start_period: 30s
networks:
- rustfs-mnmd
rustfs-node3:
image: rustfs/rustfs:latest
container_name: rustfs-node3
hostname: rustfs-node3
environment:
- RUSTFS_VOLUMES=http://rustfs-node{1...4}:9000/data/rustfs{1...4}
- RUSTFS_ADDRESS=0.0.0.0:9000
- RUSTFS_CONSOLE_ENABLE=true
- RUSTFS_CONSOLE_ADDRESS=0.0.0.0:9001
- RUSTFS_ACCESS_KEY=rustfsadmin
- RUSTFS_SECRET_KEY=rustfsadmin
- RUSTFS_CMD=rustfs
ports:
- "9020:9000" # API endpoint
- "9021:9001" # Console
volumes:
- node3-data1:/data/rustfs1
- node3-data2:/data/rustfs2
- node3-data3:/data/rustfs3
- node3-data4:/data/rustfs4
command: [ "sh", "-c", "sleep 3 && rustfs" ]
healthcheck:
test:
[
"CMD",
"sh", "-c",
"curl -f http://localhost:9000/health && curl -f http://localhost:9001/health"
]
interval: 10s
timeout: 5s
retries: 3
start_period: 30s
networks:
- rustfs-mnmd
rustfs-node4:
image: rustfs/rustfs:latest
container_name: rustfs-node4
hostname: rustfs-node4
environment:
- RUSTFS_VOLUMES=http://rustfs-node{1...4}:9000/data/rustfs{1...4}
- RUSTFS_ADDRESS=0.0.0.0:9000
- RUSTFS_CONSOLE_ENABLE=true
- RUSTFS_CONSOLE_ADDRESS=0.0.0.0:9001
- RUSTFS_ACCESS_KEY=rustfsadmin
- RUSTFS_SECRET_KEY=rustfsadmin
- RUSTFS_CMD=rustfs
ports:
- "9030:9000" # API endpoint
- "9031:9001" # Console
volumes:
- node4-data1:/data/rustfs1
- node4-data2:/data/rustfs2
- node4-data3:/data/rustfs3
- node4-data4:/data/rustfs4
command: [ "sh", "-c", "sleep 3 && rustfs" ]
healthcheck:
test:
[
"CMD",
"sh", "-c",
"curl -f http://localhost:9000/health && curl -f http://localhost:9001/health"
]
interval: 10s
timeout: 5s
retries: 3
start_period: 30s
networks:
- rustfs-mnmd
networks:
rustfs-mnmd:
driver: bridge
volumes:
node1-data1:
node1-data2:
node1-data3:
node1-data4:
node2-data1:
node2-data2:
node2-data3:
node2-data4:
node3-data1:
node3-data2:
node3-data3:
node3-data4:
node4-data1:
node4-data2:
node4-data3:
node4-data4:

View File

@@ -0,0 +1,172 @@
#!/bin/bash
# Copyright 2024 RustFS Team
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# test-deployment.sh - Quick test script for MNMD deployment
# Usage: ./test-deployment.sh
set -e
# Colors for output
GREEN='\033[0;32m'
RED='\033[0;31m'
YELLOW='\033[1;33m'
NC='\033[0m' # No Color
echo "========================================="
echo "RustFS MNMD Deployment Test"
echo "========================================="
echo ""
# Test 1: Check if all containers are running
echo "Test 1: Checking container status..."
RUNNING=$(docker-compose ps | grep -c "Up" || echo "0")
if [ "$RUNNING" -eq 4 ]; then
echo -e "${GREEN}✓ All 4 containers are running${NC}"
else
echo -e "${RED}✗ Only $RUNNING/4 containers are running${NC}"
docker-compose ps
exit 1
fi
echo ""
# Test 2: Check health status
echo "Test 2: Checking health status..."
HEALTHY=0
for node in rustfs-node1 rustfs-node2 rustfs-node3 rustfs-node4; do
STATUS=$(docker inspect "$node" --format='{{.State.Health.Status}}' 2>/dev/null || echo "unknown")
if [ "$STATUS" = "healthy" ]; then
echo -e " ${GREEN}$node is healthy${NC}"
HEALTHY=$((HEALTHY + 1))
elif [ "$STATUS" = "starting" ]; then
echo -e " ${YELLOW}$node is starting (wait a moment)${NC}"
else
echo -e " ${RED}$node status: $STATUS${NC}"
fi
done
if [ "$HEALTHY" -eq 4 ]; then
echo -e "${GREEN}✓ All containers are healthy${NC}"
elif [ "$HEALTHY" -gt 0 ]; then
echo -e "${YELLOW}$HEALTHY/4 containers are healthy (some may still be starting)${NC}"
else
echo -e "${RED}✗ No containers are healthy${NC}"
exit 1
fi
echo ""
# Test 3: Check API endpoints
echo "Test 3: Testing API endpoints..."
PORTS=(9000 9010 9020 9030)
API_SUCCESS=0
for port in "${PORTS[@]}"; do
if curl -sf http://localhost:${port}/health >/dev/null 2>&1; then
echo -e " ${GREEN}✓ API on port $port is responding${NC}"
API_SUCCESS=$((API_SUCCESS + 1))
else
echo -e " ${RED}✗ API on port $port is not responding${NC}"
fi
done
if [ "$API_SUCCESS" -eq 4 ]; then
echo -e "${GREEN}✓ All API endpoints are working${NC}"
else
echo -e "${YELLOW}$API_SUCCESS/4 API endpoints are working${NC}"
fi
echo ""
# Test 4: Check Console endpoints
echo "Test 4: Testing Console endpoints..."
CONSOLE_PORTS=(9001 9011 9021 9031)
CONSOLE_SUCCESS=0
for port in "${CONSOLE_PORTS[@]}"; do
if curl -sf http://localhost:${port}/health >/dev/null 2>&1; then
echo -e " ${GREEN}✓ Console on port $port is responding${NC}"
CONSOLE_SUCCESS=$((CONSOLE_SUCCESS + 1))
else
echo -e " ${RED}✗ Console on port $port is not responding${NC}"
fi
done
if [ "$CONSOLE_SUCCESS" -eq 4 ]; then
echo -e "${GREEN}✓ All Console endpoints are working${NC}"
else
echo -e "${YELLOW}$CONSOLE_SUCCESS/4 Console endpoints are working${NC}"
fi
echo ""
# Test 5: Check inter-node connectivity
echo "Test 5: Testing inter-node connectivity..."
CONN_SUCCESS=0
for node in rustfs-node2 rustfs-node3 rustfs-node4; do
if docker exec rustfs-node1 nc -z "$node" 9000 2>/dev/null; then
echo -e " ${GREEN}✓ node1 → $node connection OK${NC}"
CONN_SUCCESS=$((CONN_SUCCESS + 1))
else
echo -e " ${RED}✗ node1 → $node connection failed${NC}"
fi
done
if [ "$CONN_SUCCESS" -eq 3 ]; then
echo -e "${GREEN}✓ All inter-node connections are working${NC}"
else
echo -e "${YELLOW}$CONN_SUCCESS/3 inter-node connections are working${NC}"
fi
echo ""
# Test 6: Verify data directories
echo "Test 6: Verifying data directories..."
DIR_SUCCESS=0
for i in {1..4}; do
if docker exec rustfs-node1 test -d "/data/rustfs${i}"; then
DIR_SUCCESS=$((DIR_SUCCESS + 1))
else
echo -e " ${RED}✗ /data/rustfs${i} not found in node1${NC}"
fi
done
if [ "$DIR_SUCCESS" -eq 4 ]; then
echo -e "${GREEN}✓ All data directories exist${NC}"
else
echo -e "${RED}✗ Only $DIR_SUCCESS/4 data directories exist${NC}"
fi
echo ""
# Summary
echo "========================================="
echo "Test Summary"
echo "========================================="
echo "Containers running: $RUNNING/4"
echo "Healthy containers: $HEALTHY/4"
echo "API endpoints: $API_SUCCESS/4"
echo "Console endpoints: $CONSOLE_SUCCESS/4"
echo "Inter-node connections: $CONN_SUCCESS/3"
echo "Data directories: $DIR_SUCCESS/4"
echo ""
TOTAL=$((RUNNING + HEALTHY + API_SUCCESS + CONSOLE_SUCCESS + CONN_SUCCESS + DIR_SUCCESS))
MAX_SCORE=23
if [ "$TOTAL" -eq "$MAX_SCORE" ]; then
echo -e "${GREEN}✓ All tests passed! Deployment is working correctly.${NC}"
exit 0
elif [ "$TOTAL" -ge 20 ]; then
echo -e "${YELLOW}⚠ Most tests passed. Some components may still be starting up.${NC}"
echo " Try running this script again in a few moments."
exit 0
else
echo -e "${RED}✗ Some tests failed. Check the output above and logs for details.${NC}"
echo " Run 'docker-compose logs' for more information."
exit 1
fi